One type of speaker which is actually pretty common but which gets little press / attention here on audiogon is the 2.5 way.
A 2.5 way speaker is almost a 3-way, but it isn't. It is a speaker with 3 drivers, but instead of a tweeter, midrange and woofer (TMW) it lacks a true midrange. The "midrange" is really a mid-woofer, that shares bass duties with the woofer. Often these two drivers are identical, though in the Focal Profile 918 the midwoofer and woofer were actually different drivers with the same nominal diameter (6").
The Monitor Audio 200 is a current example of the concept, but I am sure there are many others. It's also quite popular in kit form. One of the most high-end kits I know of is the Ophelia based on a ScanSpeak Be tweeter and 6" Revelator mid-woofers. I haven't heard them, but I am in eternal love with those mid-woofers. I believe the original plans come from the German speaker building magazine Klan Ton.
However many other kits are also available
But regardless of kit, or store purchased, are you a 2.5 way fan? Why or why not?
@microlab If you properly account for the resistance of a coil and use quality parts, I’ve not seen a problem with crossing at 350 hz. A simple cap will typically run between 50 and 150 mic’s. An inductor can certainly add phasing problems... using an inductor and porting can cause real phase issues in a woofer, but with proper crossover point selection, using driver compensation and quality parts, most issues can be avoided. I would highly recommend that when you can to just avoid using 12/12 slopes, that will normally stop the need of having to reverse polarity. 6/12 or 12/18 butterworth are both more phase coherent than 12/12 anyway. If you need to run 12/12, choose linkwitz slopes.
As a DIY speaker builder, I've logged many hours building & tweaking crossovers, and one big compromise of a 3-way design is the HP filter you have to put on the midrange. This is, at the very least a large capacitor to roll off at ~350 Hz, plus a large inductor too for a 2nd order. These components are now in the signal path of your crucial midrange frequencies, and they're also affecting the phase alignment of the midrange with the tweeter. You may also need to invert the polarity of the mid driver (or not)... it just depends on other parts of the design. The 2.5-way design has fewer crossover components in the way of the midrange, and has more low end capability than just a 2-way. There still will be some phasey issues between the 2 woofers as the lower woofer rolls in to align with the upper woofer. I admire the idea, but I've never built or owned a 2.5 way design.
Well I am a huge 2.5 way fan. My fav are the Wilson Benesch. As suggested they use the same size midrange as woofer and work very hard to create a woofer that is as fast as the mid range so there is no loss of coherence. The woofer does woof and the mid because it is 7" can go down to 500hz. I think unlike many 2.5 way speakers the WB's midrange goes all the way up to 5000 kHz and this covers almost the entire midrange with male voices and cellos sounding very full and rich.
Very good post , helped to clarify things I kinda , sorta knew . I'll give you a joke I just heard on the local jazz station . Announcer was interviewing a local retired trumpet player who still does some sub and gigs around the Twin Cities . The dude had once played for Lawrence Welk , both of them rolled off the names of some pretty good musicians who played for the bubble man over the years . DJ kinda beat around the bush as to why so many good players played the corn ball . Trumpet man said , "you have to be a hell of a musician to play that bad all the time " !
This has been a great thread for me to understand a little bit better about something I enjoy. Thanks Eric for the thread and Tim with your long term experience in speaker building for all your supporting information.
Most of the human voice in singing operates from 80 for very low bass male voice to around 1500 which is above natural female vocals to maybe someone trying to hit an extreme note to impress. Everything that Erik has quoted throughout this thread is correct. A couple of things that I thought of reading through the thread. **Crossing a tweeter low: sometimes its a no brainer, the mid has a dip or peak and you need the tweeter to cover that range for a smooth frequency transfer... Other times, I have listened to the mid and the tweeter both at a given frequency and thought the tweeter sounded better and chose to cross lower for that reason. The rule of thumb for how low you can cross a tweeter is 2x its resonance at 12db per octave. This can be broken depending on tweeter, but it is a good general rule to follow. For me an optimal in theory speaker would be a 3 way crossed somewhere around 80 on bottom and 2200 to 3k on top. This keeps the deep bass off the mid which really helps keep it clean and it also keeps the crossover out of any critical vocal region. Of course finding perfect parts to put into practice are not always so easy. Erik, you’ve done a great job with this thread. Getting folks to understand a 2.5 way vs a 3 way isn’t always easy. Tim
I can't disagree with you in terms of the voice, but I will say, I just haven't found the upper crossover point to be a really clear indicator of performance either way.
I think the bottom line is that if a driver is crossed over close to 4kHz, it’s covering most of a human’s vocal capability that falls within the most sensitive range of our hearing ability. It keeps the crossover away from the meat of the music. Even if harmonics do extend as far as 17kHz, many folks have trouble hearing freqs beyond 8kHz.
It’s been a while since I posted but this discussion is good. In 2004 I bought on an Audiogon auction the Soliloquy 6.3 2.5 way speakers designed by Phil Jones. An excellent review can be found here including crossover specs: https://www.thefreelibrary.com/Soliloquy+Model+6.3+loudspeakers-a097823139 I’ve changed every component of my rig, currently Butler TDB2250, Krell KBL and Denon DVD-3910 and the Soliloquy’s have allowed me to hear the changes. I vote that when done right the 2.5 ways are wonderful.
I am very happy with my Advent Heritage speakers which are a 2.5 way design. This is for two main reasons - they were a little over a hundred dollars including the refoam kit and they sound great. Also they come in a cabinet that is solid pecan front and back. Also they are fairly efficient. Lovely bass extension and nicely detailed treble. It did take several hours to replace the surrounds on all 4 drivers but so worth it. Currently powering them with Kenwood M1/C1 set at 105 wpc.
Erik, they stated "fundamental" . For harmonics they claim 17kHz (hard to believe).
I found that high "C" for soprano is about 1kHz, but they can produce "whistle tone" (or falsetto) at D7 or 2349.3 Hz and this is the highest voice human being can produce - hence 3kHz is not even possible.
Erik, Paradigm Studio 60 v2, I had before, was sold at about $1k. It usually means that manufacturer sell them for $500. In my company cost of materials is about 30% of the cost of the product. That would imply $170-200 for three speakers, box, xover etc. I did not like sound of the metal dome tweeter and found it to be one of the cheapest Vifa tweeters available. It is very tempting to build speaker using the best drivers available, including underhung motors etc. Unfortunately I already attempted to replace this Vifa tweeter and to redesign xover with high quality components. The purchase of Hyperion speakers was direct result of this attempt. Perhaps I'll learn more and try again one day.
Be careful reading that. There is a significant difference between the fundamental and the range. The fundamentals are the bottom, pure note. But the range is far higher.
that states: 350Hz - 3kHz for fundamental frequency - about 15 times higher. I don't doubt Wikipedia and just wonder. Could this be that "typical" voice is in narrower range while soprano can get 15x (almost 3 octave) higher? Perhaps it is "talking" (Wikipedia stated "speech") vs. "singing".
I appreciate your comment about "not laser like" start of beaming phenomena - my tweeters might already provide wider dispersion where large midrange speaker starts to beam.
@kijanki - This is probably not for you, but here is a very different type of speaker that has a lot of fans. The Seas A26. 2-way with a 10" mid-woofer and single capacitor crossover (6db/Octave):
2. Isn’t crossing below 2kHz to close to resonance of most tweeters? I’ve read recommendation to keep crossover frequency at least 1-2 octaves above tweeter’s resonance.
It's at the bottom end of most tweeters, yes. Usually the rule of thumb is to be 2-3x the resonance. What really matters though is not the resonance but distortion and excursion. That is, overwhelming the tweeter. So your choice in tweeter, as well as the slope and knee point matter. The ring-radiator I use in my desktop speakers is the "deep chamber" version with significantly lower usable crossover point than the normal variety. The Mundorf AMT I use in my living room has awesome power handling, distortion and a relatively low resonance point.
The crossover also matters. A tweeter may be able to safely be crossed at 2kHz using a fourth order (24 dB/octave) filter, but 5 kHz using a first order filter (6 dB/octave). You see this with the Joseph Audio speakers discussed in other threads. With a fast drop off they push the tweeter lower than average.
3. Break-up modes for typical woofer might be around 5kHz. Isn’t cutting at 2kHz too close - especially for 6dB/octave? In comparison my 3-way speakers cut at 230Hz and 3kHz - outside of "sensitive" zone.
This really depends on the driver and crossover, and what exactly you mean by a "woofer" or "mid-woofer." The combination of slope, and filters used in the woofer may help push the usable range up.
Also, I agree with you that small sweet spot, caused by beaming, might be advantage in acoustically bad rooms, but that’s only if you listen alone. For me wide sweet spot is very important. If I’m not mistaken beaming for 6.5" drivers starts around 1.5kHz.
Even if a 6.5" starts beaming, it is not like a LASER. It doesn't switch from omni-directional to flash light at 1.5 kHz. It just gets narrower as it goes up. If your tweeter's dispersion matches you will still have very nice sound off axis, as the entire speaker will seem to diminish, and not just one particular range. I do not think that a 6" mid-woofer in a 2 or 2.5 way is a terrible idea at all, or that you will have a pin-point sweet spot, but it can give you added clarity when you are constrained in where you put them.
There are many designers who push the idea that crossovers are bad. Either you want no crossover, or 1st order, or you want the crossover completely out of the vocal range, etc. I'm not really with them. Personally ( and I do not insist that you agree with me ) I have not heard a problem with a well implemented crossover in the 1-3kHz range. I have also not fallen in love with Thiel or Vandersteen or any other perfect time aligned speaker.
I'm happy to have learned that at least on this subject, Joseph Audio agrees with me, and none of their fans mention any sort of discontinuity in the vocal ranges.
1. Isn’t crossing at 2kHz or lower in the range of female voice, that has fundamentals between 350Hz and 3kHz?
2. Isn’t crossing below 2kHz to close to resonance of most tweeters? I’ve read recommendation to keep crossover frequency at least 1-2 octaves above tweeter’s resonance.
3. Break-up modes for typical woofer might be around 5kHz. Isn’t cutting at 2kHz too close - especially for 6dB/octave? In comparison my 3-way speakers cut at 230Hz and 3kHz - outside of "sensitive" zone.
Also, I agree with you that small sweet spot, caused by beaming, might be advantage in acoustically bad rooms, but that’s only if you listen alone. For me wide sweet spot is very important. If I’m not mistaken beaming for 6.5" drivers starts around 1.5kHz. Crossing at 2kHz or lower might help but you need tweeter with very low resonance frequency. I’m not sure what it is, since I don’t design speakers but suspect that it is around 1kHz. My speakers should beam (6.5" midrange) but they have wide sweet spot. Perhaps weird design of the midrange makes a difference (ferrofluid instead of spider web suspension and the wide flat disk instead of dust cap)
The Spendor’s also pair quite well with Linn and Naim gear, which my friends and I all own. It’s a classic combination, which certainly make a difference, along with the room.
Although I own 30 year old, 3 way Thiel 3 way speakers (CS2) and still enjoy them very much, when friends have asked me what speakers to buy in recent years, I recommended Spendor...the A5 (now discontinued) and the D7 (current). Both are 2.5 way speakers and both owners love them. I do too. What’s not to love?
At the end of the day, it’s about what sounds good to you, regardless of speaker design or price (assuming you can afford it?)
I tend to crossover at 2 kHz or slightly lower since I have only really made 2-ways so far, and don't have an issue with any coherency. One pair sits on my desktop. The near field listening is a real treat given my living room is an acoustic mess.
The only 3-ways I currently own are Klipsch Heresy IIIs. I just looked up their crossover point, it's a rather high 5kHz. These speakers still lack coherency unless listening from a distance of 11' or more. I would guess that is most likely due to horn beaming, though the others I mentioned also benefit from such distance. I sit about 9' from the speakers in my main rig. Of all the 3-ways I've tried, none work for me in such proximity. I imagine a design like the Elac Adantes might work.
Interesting! I always assumed otherwise. Especially with the Monitor Audio, that tweeter doesn't really have a lot of low end extension.
Definitely not how I would do a 3-way at all, but then hey, I'm not making the big bucks designing speakers. It'd get a very lively, composite midrange and cross it over at 3-5kHz
Usually higher tweeter points are achieved with 3-way systems which use a "true" midrange.
That's why I stated, "well designed examples."
I know of quite a few 3-ways that are still crossed over to the tweeter rather low, often in the 2kHz range. Some 2-ways are crossed this low, but it often doesn't work so well IMO. There's no hard rules in any of this, but I rarely come across 3-ways that have the coherency of well designed 2-ways.
Yep, the Wilson Benesch Vector is a funky beast indeed!
Not only are they using no crossover on the midrange, they are using 1st order slopes on the tweeter and woofer, but wait! There’s more!
They are also porting the mid-range unit. Not unheard of, but rather rare. Only other speaker I can think of with similar dual-porting is from Sonus Faber.
I don’t think calling it a 2.5 way really does it justice since it really has only 2 filter sections, which would normally be for a 2-way speaker. It’s pretty miraculous they are able to do all of this and still maintain a comfortable impedance that they do.
In some ways, this reminds me of the famous Seas A25 speaker, now available as the A26 kit speakers.
I have a pair of Wilson Benesch Vectors. My understanding of their 2.5 way system is 500hz for the woofer, 5khz for the tweeter, and a full range “midrange”. Seems to work well.
I love my ProAc Studio 148, which I think qualify...
"The two 6.5 inch units cover bass frequencies but each one has been mechanically tuned to cover mid frequencies. To improve dispersion the upper bass driver has a phase plug whilst the lower unit has a ProAc open weave cap."
So, just idle curiosity,
but can a 2.5way have with multiple woofers--that is at least four or more total drivers--or is that a different beast all together?
@kijanki Sounds to me like you've got a big thing for D' Apolito areays. I'd hardly say those lack their own technical challenges. They're even more directionality challenged since the wave fronts decouple based on the distances of the centers of the drivers instead of the driver diameter itself. A 2.5 way design can be realized with 6.5 inch mid-bass and woofers and not suffer the problems you claim through appropriate alignment, crossover slopes, and polarity.
Hmm...seems no one pointed out the real advantages of 2.5-ways (at least well designed examples). It's the same advantage of 2-ways: the high crossover point and resulting coherency. In a well designed 2-way, it's often difficult to make out the separation of drivers, whereas I rarely encounter a 3-way that achieves the same feat. I can usually hear that there's 3 or more drivers doing the work.
A midwoofer and tweeter crossed over at say, 3.8kHz, benefit from coherency and lack of crossover distortions where the meat of the music lies: in the midrange. This is largely why speakers like Harbeths and Spendors can produce vocals that rival some electrostats.
Of course there can be disadvantages to a high crossover point, but I find they are insignificant or inaudible in most high quality 2/2.5-ways.
Erik, I think you might be onto something. I couldn’t be happier with the sound and size of my 2 1/2 way Epos M16i’s. They sound good at both low and high volumes and with all types of music. I also currently own ML Monti’s and Harbeth Super HL5+ and as much as I really like them both, if forced to choose the Epos just might win out as it’s just a really good all purpose speaker.
What I mean to say is, there is no 1 ideal dispersion pattern. From ESL's to open baffles, to traditional multi-way speakers, some with horns. All have very different radiating patterns, and all have ardent fans.
It is not a dispersion pattern for particular speaker but for the group of them. Speakers that have large membrane (2.5 way) will suffer.
Thanks Erik. I had assumed (perhaps wrongly) that in a 2.5 way, only the woofer physically closer to the tweeter plays up to the crossover frequency of the tweeter (say 3 kHz), and the other woofer has its low pass set to a lower frequency (say, 500 hz). I thought that was done to prevent lobing, but defer to you.
In such an arrangement, the level matching seems a bit complicated because some of the frequency range up to 3 kHz is covered by two drivers and some of it by one.
Indeed, but this is what crossover designers must do anyway. In any multi-way system there is almost always some level matching that must occur, in this case that as well as the dimensions of the upper woofer enclosure must be carefully balanced.
What I mean to say is, there is no 1 ideal dispersion pattern. From ESL's to open baffles, to traditional multi-way speakers, some with horns. All have very different radiating patterns, and all have ardent fans.
Beaming is not, by itself, bad. It may not be ideal for you in your listening room however.
For the record, my current speakers are traditional 2-way designs, with subwoofer for music. I still like 2.5 ways though. :)
Are there challenges in designing a crossover for a 2.5 way given that a large portion of the frequency range (bass) is shared by two identical drivers, but another large portion (midrange) is handled by just one driver? I was thinking in terms of level matching due to that transition from two drivers to one driver.
You say beaming is a disadvantage? I disagree! :) A speaker with tight dispersion can sound better in the sweet spot, especially with an acoustically messy room.
It all depends on your listening style and location.
Some of Magico's smaller speakers are very wide dispersion. Great stereo imaging no matter where, but the trade off is they need a lot of room or great room acoustics.
You must have a verified phone number and physical address in order to post in the Audiogon Forums. Please return to Audiogon.com and complete this step. If you have any questions please contact Support.
